pytorch/benchmark

Pull Requests Commits

Update requirements.txt

atalman committed 318 days ago

Verified 86d67336

Pin opacus to 1.5.3 on opacus_cifar10 model

atalman committed 318 days ago

Verified a32be710

update torchvision pin (#154255)

BoyuanFeng committed 318 days ago

2caf997d

Use cuBLAS default workspace size in Lt (#153556)

generatedunixname499836121 committed 319 days ago

e759c474

Fix CHROMIUM_EVENT_LOG being none

jamesjwu committed 322 days ago

4617d91c

Don't upload compiler benchmark debug info to the benchmark database (#153769)

huydhn committed 322 days ago

6693f584

fix anti-pattern for cudagraph (#2619)

BoyuanFeng committed 323 days ago

227fb01e

Allow higher fp16 tolerance for phlippe_resnet on CUDA 12.8 (#154109)

huydhn committed 323 days ago

1b4c3199

Extend empty_gpu_cache to mps (#153657)

malfet committed 329 days ago

316981ee

Fix test broken by D73809989 (#153413)

masnesral committed 331 days ago

e32ebfe2

Optimize dedupe region ancestor tracking (#152589)

mlazos committed 331 days ago

f3ef83db

Fix misleadingly high AOT Inductor dashboard performance (#153060)

generatedunixname499836121 committed 331 days ago

d36d4778

Use universal flatten APIs (#152505)

mlazos committed 331 days ago

ecf479d0

Track node mutations (#152389)

mlazos committed 331 days ago

7345ba37

Revert "Track node mutations (#152389)"

generatedunixname499836121 committed 332 days ago

99082650

Revert "Use universal flatten APIs (#152505)"

generatedunixname499836121 committed 332 days ago

595677cd

Update ruff to 0.11.8 (#153249)

generatedunixname499836121 committed 333 days ago

9e0e6b2a

Use universal flatten APIs (#152505)

mlazos committed 333 days ago

ee027de2

Track node mutations (#152389)

mlazos committed 333 days ago

5c048d19

Revert "`has_triton`: Use the device interface for detecting Triton availability (#139171)"

generatedunixname499836121 committed 335 days ago

d4a5c2aa

Fix Readme instruction (#2615)

ZainRizvi committed 337 days ago

232e23ff

Support activation quantization without scaling (#2607)

mengluy0125 committed 338 days ago

3a270fa9

`has_triton`: Use the device interface for detecting Triton availability (#139171)

generatedunixname499836121 committed 338 days ago

de736fee

Refactor the fetching of subgraph names (#152770)

anijain2305 committed 339 days ago

5838b079

Enable max autotune for AOTInductor benchmark (#149309)

exclamaforte committed 347 days ago

e53141fe

Opt-in unified cuBLAS + cuBLASLt workspaces (#151163)

generatedunixname499836121 committed 352 days ago

34f7cc03

context manager/decorator for dynamo config patching during tracing (#150586)

williamwen42 committed 352 days ago

eab7dd49

Put "everything" WaitCounters in dynamo_timed (#151757)

masnesral committed 352 days ago

2713258b

Fix duration logging for dynamo_compile (#151749)

masnesral committed 352 days ago

3dd99cb3

Handling of skip category of tests (#2597)

arathi-hlab committed 353 days ago

92a05f1c

Older